| 00:20.13 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/src/libbu/semaphore.c: should be fine to use bu_bomb instead of abort as bu_bomb should not attempt to acquire a semaphore (it doesn't use bu_log). |
| 01:11.56 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/src/libbu/bomb.c: add comment docs on bu_bomb usage and add name given extensive changes over time. |
| 01:16.49 | *** join/#brlcad akreal (n=ak@87.249.56.198) | |
| 02:13.01 | *** join/#brlcad CIA-4 (n=CIA@208.69.182.149) | |
| 02:54.33 | *** join/#brlcad MinstrelGypsy (n=mario_du@bas2-sudbury98-1177871953.dsl.bell.ca) | |
| 02:54.36 | *** join/#brlcad ak__ (n=ak@87.249.56.198) | |
| 03:02.41 | MinstrelGypsy | http://irix32.spaces.live.com/photos windowsside albumn, 7.10.1 sorta working. |
| 04:13.15 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/src/other/tk/unix/tcl.m4: if we're on freebsd, there should be no reason why we can't just use gcc instead of ld -- this allows 'lib' flags being provided for XFT_LIBS via xft-config to actually not break the build. |
| 04:31.39 | *** join/#brlcad louipc_ (n=louipc@bas8-toronto63-1177614282.dsl.bell.ca) | |
| 04:55.40 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/solids.sh: set the ld_library_path so the binaries invoked by the test can find the tcl/tk libraries. also, write out the mged script to a file so that errors don't dump the entire script. use solids.log for the log. |
| 05:07.33 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1088754495.dsl.bell.ca) | |
| 05:13.39 | jack- | guys, question..is adrt more mature meanwhile? like..should i include it in a brlcad package? |
| 05:13.58 | jack- | (too lazy to study changelogs, sorry!) |
| 05:14.29 | jack- | it's wicked on macs, anyway |
| 05:14.40 | jack- | since SDL is aqua (apple windowmanager) stuff |
| 05:14.47 | jack- | and the rest is x11 stuff |
| 05:15.17 | jack- | so its not too easy to get a smooth env while working, kind of |
| 05:25.49 | brlcad | jack-: probably not |
| 05:26.02 | jack- | better leave adrt out again? |
| 05:26.07 | brlcad | yes |
| 05:26.13 | jack- | ok, thx :) |
| 05:26.22 | brlcad | it's not been fleshed out build-wise with any sort of polish |
| 05:26.41 | jack- | thats np, i mean i could dig minor build issues myself |
| 05:26.51 | jack- | but for end users, as a frontend? |
| 05:27.01 | brlcad | it'll build on os x, but it requires a fair bit of manual work |
| 05:27.19 | jack- | i know the sdl tricks pretty well i guess :) |
| 05:27.19 | brlcad | and even once you have it up, there is practically no documentation |
| 05:27.26 | brlcad | it's not meant for end-user use just yet |
| 05:27.26 | jack- | ok |
| 05:27.33 | jack- | alright, no adrt then |
| 05:29.33 | jack- | brlcad: no clue if you're the one who tried to build/use it on mac os x |
| 05:29.47 | jack- | but if you want a much easier life for such things, just get fink |
| 05:29.58 | jack- | saved me so much time+headaches already :) |
| 05:33.14 | jack- | brlcad will appear in fink soonish, i'm just finishing my package |
| 05:33.54 | jack- | (http://pdb.finkproject.org/pdb/maintainer.php?maintainer=jackfink = me) |
| 05:36.13 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/Makefile.am: mged_solids.log is no longer generated |
| 05:38.24 | brlcad | er, i've never "tried" .. i build it when i want to use it by hand ;) |
| 05:38.41 | jack- | :) |
| 05:38.57 | brlcad | glad to hear someone's picking up the fink package for it |
| 05:39.00 | jack- | building anything you know on a new system could always be seen as a try |
| 05:39.10 | jack- | nothing wrong with that :) |
| 05:39.25 | brlcad | i would, but me and the fink devs differ on many issues that I just try to avoid the conflict |
| 05:39.34 | jack- | oh |
| 05:39.38 | jack- | like what? |
| 05:40.14 | jack- | in my eyes, the dp/mp folks are way more of elitist whackos you can have troubles communicating with |
| 05:40.20 | jack- | hence fink |
| 05:40.50 | brlcad | nothing important |
| 05:42.05 | jack- | i like discovering and packaging great things like brlcad |
| 05:42.31 | jack- | sphinx is kind of similar, an awesome opensourced speech recognizing system done by edu folks |
| 05:43.22 | jack- | only stuff holding that one back was some odd g95 thing, it just wont build if you have fortran installed |
| 05:43.53 | jack- | unfortunately a system component, outside of fink ;) i should hack the autoconf files a bit |
| 05:44.05 | brlcad | yeah, sounds like something in a configure.ac |
| 05:44.14 | jack- | yup |
| 05:45.12 | jack- | sdl is fairly neat once you get used to shit on *.framework and use only the fink stuff |
| 05:45.29 | jack- | some things like sdl, mysql and so on are available from many sources for macs |
| 05:45.36 | jack- | wicked..but you can handle that |
| 05:53.36 | brlcad | sdl from the sdl folks works well enough too -- but requires a fair bit of build-system mods to support them regardless (at least portably) |
| 05:57.38 | jack- | yup |
| 05:57.58 | jack- | or a fair bit of packager brainwork to adapt stuff |
| 05:58.27 | jack- | i dont like all the purely mac-ish things like frameworks, the unix way is so much easier |
| 05:58.39 | jack- | and not any less reliable if you know what you're doing |
| 06:03.05 | jack- | easier = more portable..thanks to fink i only have to make sure stuff uses the right include- and libdirs |
| 06:03.40 | jack- | and autoconf/make-upstream is influenced enough to make sure it works |
| 06:03.54 | jack- | so fink is really comfortable, kind of |
| 06:04.08 | brlcad | is that a soapbox you're standing on? :) |
| 06:04.18 | brlcad | to each their own ;) |
| 06:04.27 | jack- | yeah |
| 06:04.57 | jack- | stuff is getting harder now with intel vs ppc anyway |
| 06:05.15 | jack- | cant build fat binaries and libs with fink yet :) but its ok |
| 06:07.32 | brlcad | did they finally add smp build support? |
| 06:08.23 | jack- | not directly through fink, but some even some core guys do it |
| 06:08.40 | jack- | np to get -j4 into your makeflags if you know what you're doing |
| 06:09.02 | brlcad | i know it's doable, I made the mods for years |
| 06:09.12 | jack- | some stuff wont work, but most stuff will |
| 06:09.28 | jack- | its just not ripe enough to include it in fink for "users" |
| 06:09.51 | jack- | (people who want to build things without being able to do the same stuff "manually" from upstream tarballs or so) |
| 06:33.34 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/ (7 files): set the ld_library_path so the binaries invoked by the test can find the tcl/tk libraries. |
| 06:39.54 | *** join/#brlcad dtidrow_work (n=dtidrow@host169.objectsciences.com) | |
| 07:42.27 | *** join/#brlcad ak__ (n=ak@87.249.56.198) | |
| 08:07.54 | *** join/#brlcad ak__ (n=ak@87.249.56.198) | |
| 09:10.51 | *** join/#brlcad akreal (n=ak@87.249.56.198) | |
| 09:28.09 | *** join/#brlcad louipc_ (n=louipc@bas8-toronto63-1177613669.dsl.bell.ca) | |
| 10:24.30 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-094-012.pools.arcor-ip.net) | |
| 10:52.03 | *** join/#brlcad Elperion (n=Bary@p54876A32.dip.t-dialin.net) | |
| 11:13.02 | *** join/#brlcad akreal (n=ak@87.249.56.198) | |
| 11:18.42 | *** join/#brlcad Waimea (i=clock@77-56-101-79.dclient.hispeed.ch) | |
| 12:13.44 | *** join/#brlcad elite01_ (n=elite01@dslc-082-082-067-161.pools.arcor-ip.net) | |
| 12:36.30 | *** join/#brlcad thing0 (i=thing0@124-168-78-7.dyn.iinet.net.au) | |
| 12:36.38 | thing0 | hey guys |
| 12:36.42 | thing0 | long time no see |
| 12:36.51 | thing0 | haven't been here in years :) |
| 13:35.38 | *** join/#brlcad tarzeau (i=sengun@krum.ethz.ch) | |
| 13:43.43 | *** join/#brlcad poolio (n=poolio@c-69-251-3-107.hsd1.md.comcast.net) | |
| 14:03.44 | *** join/#brlcad ak__ (n=ak@87.249.56.198) | |
| 14:16.44 | bjorkBSD | has it changed much, thing0? |
| 14:17.02 | tarzeau | is it possible make clean target is not very good? |
| 14:17.07 | tarzeau | like it removes stuff that's needed? |
| 14:21.21 | thing0 | umm |
| 14:21.26 | thing0 | yeah |
| 14:21.30 | thing0 | there is more people ;) |
| 14:21.34 | thing0 | i've changed too |
| 14:21.35 | thing0 | hehe |
| 14:26.38 | thing0 | so how is BRLCAD actually going? |
| 14:33.41 | ``Erik | a little to the left |
| 14:48.51 | thing0 | lol |
| 15:19.34 | tarzeau | itclStubLib.lo: In function `Itcl_InitStubs': |
| 15:19.35 | tarzeau | /var/www/debian/brl-cad/brlcad-7.10.0/src/other/incrTcl/itcl/generic/itclStubLib.c:71: undefined reference to `tclStubsPtr' |
| 15:19.40 | tarzeau | when i try to build it (on debian gnu/linux) |
| 15:19.57 | brlcad | hello thing0 |
| 15:20.32 | thing0 | hi brlcad |
| 15:20.36 | thing0 | long time no see |
| 15:20.43 | brlcad | tarzeau: can you try to build latest CVS instead of .0 |
| 15:20.49 | thing0 | i've been cadding it up lately |
| 15:20.50 | thing0 | :P |
| 15:20.56 | tarzeau | brlcad: sure. let me get that |
| 15:21.00 | brlcad | the build system has been in rapid flux wrt tcl/tk/itcl/itk |
| 15:21.00 | thing0 | actually got a job as a CAD teacher |
| 15:21.01 | thing0 | hehe |
| 15:21.10 | brlcad | thing0: nifty |
| 15:21.22 | thing0 | yeah |
| 15:21.24 | thing0 | but I quite |
| 15:21.26 | thing0 | quite |
| 15:21.28 | tarzeau | any of you use blender, wings3d, misfit model 3d and/or www.sauerbraten.org ? |
| 15:21.30 | thing0 | last try |
| 15:21.31 | thing0 | quit |
| 15:21.32 | thing0 | hehe |
| 15:21.47 | brlcad | tarzeau: i've used at least two of those |
| 15:21.49 | thing0 | tarzeau: I have used blender |
| 15:22.02 | brlcad | the first two for me |
| 15:22.14 | brlcad | none of those being a solid modeler or cad modeler too ;) |
| 15:22.32 | tarzeau | yeah i know it's different stuff :) |
| 15:22.39 | tarzeau | that's why i try to package brlcad (again) for debian |
| 15:22.48 | tarzeau | that ronja friend's using brlcad |
| 15:22.59 | tarzeau | http://ronja.twibright.com/ |
| 15:24.01 | brlcad | he pops in from time to time |
| 15:24.51 | tarzeau | clock-? |
| 15:26.34 | brlcad | yeah |
| 15:27.19 | brlcad | unless that's you and I'm getting nicks confused |
| 15:28.08 | tarzeau | no no that's not me |
| 15:28.18 | tarzeau | you're absolutely right |
| 15:28.31 | tarzeau | but i was here too, i think a year or more ago |
| 15:28.54 | tarzeau | trying to build the cvs version now |
| 15:29.36 | thing0 | how many devs are working on BRLCAD? |
| 15:31.10 | brlcad | thing0: oh it varies in any given week -- but about 4 or 5 regularly active throughout this past year |
| 15:32.05 | brlcad | the magnitude of activity also fluctuates heavily, as does peripheral support for solving things like mathematics and algorithms issues |
| 15:32.43 | thing0 | ahh ok |
| 15:33.10 | brlcad | one guy working on FEA integration, another on build system and image management refactoring, another implementing the brep support, etc |
| 15:37.28 | thing0 | ahh ic |
| 15:37.37 | thing0 | is there a mechanical simulation? |
| 15:37.48 | thing0 | as in gravity, force analysis etc. |
| 15:38.40 | brlcad | no no |
| 15:39.08 | brlcad | that's quite outside our scope at least at the moment -- that's partly why, though, there are hooks to FEA being put in place |
| 15:40.32 | thing0 | hmm |
| 15:41.05 | thing0 | should join up with that free physic engine group |
| 15:41.52 | brlcad | you need a simulation environment for a physic engine to make sense ;) |
| 15:42.21 | thing0 | yeah |
| 15:42.24 | thing0 | exactly |
| 15:43.08 | thing0 | hmm |
| 15:43.33 | thing0 | is BRLCAD actually parametric? |
| 15:43.38 | thing0 | I cannot remember |
| 15:43.45 | thing0 | I am on a backup dialup account atm |
| 15:43.51 | thing0 | so the PDF is really slow to load |
| 15:44.00 | brlcad | no, it's not parametric |
| 15:44.09 | brlcad | though we're moving towards adding that style geometric support |
| 15:44.35 | brlcad | brl-cad supports multiple representation formats, but _prefers_ implicit CSG constructions |
| 15:45.42 | brlcad | that said, we're in the process of implementing brep support beyond the existing support for meshes so that you can define parametric surfaces |
| 15:46.16 | brlcad | that's massive effort, though, and going to require a bit shift in the editing side as well as the foundation geometric engine aspects |
| 15:47.17 | brlcad | jason has almost got the base brep support implemented in terms of ray-tracing, quite awesome progress actually |
| 15:47.50 | brlcad | from there we'll need to implement brep tessellation, brep on brep evaluation, and csg to brep translation |
| 15:47.52 | tarzeau | still /var/www/debian/brl-cad/brlcad-0+20070603/src/other/incrTcl/itcl/generic/itclStubLib.c:71: undefined reference to `tclStubsPtr' |
| 15:47.57 | thing0 | ll |
| 15:47.59 | thing0 | well |
| 15:48.06 | thing0 | if I can request a feature |
| 15:48.19 | thing0 | please make the solids themselves parametric |
| 15:48.24 | thing0 | as in |
| 15:48.30 | thing0 | the primative shape |
| 15:48.31 | thing0 | s |
| 15:48.37 | thing0 | can be deformed |
| 15:48.44 | brlcad | tarzeau: can you pastebin the configure summary? |
| 15:48.45 | thing0 | no need to make base sketches and then extrude |
| 15:48.56 | brlcad | it's in your config.log near the bottom if you don't have it in a buffer |
| 15:49.02 | *** join/#brlcad smallfoot- (i=smallfoo@clamwin/translator/smallfoot) | |
| 15:49.03 | thing0 | this annoys me greatly in the current group of parametric CADs |
| 15:49.07 | tarzeau | brlcad: sure: http://gnu.ethz.ch/debian/brl-cad/brlcad_0+20070603-1_i386.build |
| 15:49.13 | brlcad | thing0: that's all planned |
| 15:49.16 | tarzeau | 500kb, very bottom |
| 15:49.37 | thing0 | thank you brlcad |
| 15:49.49 | brlcad | if you create a sphere, it'll make just an implicit sphere -- but still let you grab a point on that sphere and pull/poke it (at which point it's automatically turned into a brep spline surface) |
| 15:49.50 | thing0 | I have yet to look at onespaces offering |
| 15:49.59 | thing0 | ahh nice |
| 15:50.07 | thing0 | what I would like is that |
| 15:50.29 | thing0 | if for example I made the mid of a sphere join to the face of a cube |
| 15:50.48 | thing0 | I could define the length of the sphere - cube with a parameter |
| 15:51.02 | thing0 | that is mapped to the quadrant of the sphere |
| 15:51.08 | thing0 | and the end of the cube |
| 15:51.49 | brlcad | sounds like parametric constraints |
| 15:51.57 | thing0 | yeah |
| 15:52.03 | thing0 | but in a different way |
| 15:52.25 | thing0 | most CADs would have this as a reference |
| 15:52.33 | thing0 | I want this as a driving constraint |
| 15:52.45 | thing0 | sure I could make this shape with a base sketch |
| 15:52.51 | thing0 | but that's what I don't want |
| 15:53.06 | thing0 | cause if someone picks a different spot to make a base sketch |
| 15:53.12 | thing0 | then the way your thinking |
| 15:53.18 | thing0 | it becomes ANNOYING |
| 15:54.31 | thing0 | yeah |
| 15:55.00 | thing0 | we work in 3d but still use 2d methods to make our 3d objects |
| 15:59.02 | brlcad | we, who's we? :) |
| 16:00.12 | brlcad | brl-cad never has supported 2d methods very well, or at all for the longest time -- it was a long time before sketches were added and they're still a bastard cousin implementation that are discouraged |
| 16:00.47 | thing0 | hehe |
| 16:00.50 | thing0 | sorry brlcad |
| 16:01.02 | thing0 | just thinking of commerical parametric CAD |
| 16:01.12 | thing0 | had the sales doctorine |
| 16:01.17 | thing0 | its in my blood now |
| 16:01.18 | thing0 | :P |
| 16:01.42 | brlcad | autocad is another story ;) |
| 16:02.47 | thing0 | ahh yes |
| 16:02.50 | thing0 | autocad |
| 16:03.06 | thing0 | now with solids that have histoy |
| 16:03.10 | thing0 | *history |
| 16:10.54 | thing0 | I was teaching Autodesk Inventor |
| 16:11.01 | thing0 | but yeah |
| 16:11.17 | thing0 | with AutoCAD coming out with that 3d stuff in 2007 it was a bit weird |
| 16:12.01 | brlcad | autocad will probably forever be seen as the "2d approach" no matter how much 3d they stuff in |
| 16:12.30 | brlcad | for 3d, you grab unigraphics, pro/e, solidworks, or catia |
| 16:13.06 | brlcad | at some point down the road, BRL-CAD will be in that list as well moreso than it is today |
| 16:14.02 | thing0 | that would be good to see |
| 16:17.13 | thing0 | how long do you think it will take> |
| 16:17.13 | thing0 | ? |
| 16:17.13 | brlcad | we have most of the geometry engine aspects, particularly now with brep support being integrated |
| 16:17.13 | thing0 | like |
| 16:17.13 | thing0 | there is a bunch of free guys atm |
| 16:17.13 | brlcad | having a new enticing gui will be where the win can be made |
| 16:17.21 | thing0 | ic |
| 16:17.30 | brlcad | most users only care about the gui, not raw capabilities |
| 16:17.37 | brlcad | brl-cad has extensive raw capability |
| 16:17.40 | thing0 | hehe |
| 16:17.43 | brlcad | but a difficult gui |
| 16:17.45 | thing0 | yeah |
| 16:18.00 | thing0 | just needs to be stupid simple |
| 16:18.06 | thing0 | for the masses :P |
| 16:18.13 | brlcad | which is contrary to cad design in general |
| 16:18.28 | brlcad | cad is a complex domain, lots of different wants and expectations |
| 16:19.07 | brlcad | there's a reason there's nothing open source even remotely close to a commercial package like pro/e |
| 16:19.09 | thing0 | yeah I know |
| 16:19.40 | brlcad | aside from the 100+ cummulative staff years that they invest every single year.. |
| 16:20.10 | thing0 | yeah |
| 16:20.22 | brlcad | because it's a multi-billion dollar industry, everyone does their work in private and throws major commercial money at problems |
| 16:20.22 | thing0 | but once the product gets more brand regonition |
| 16:20.30 | thing0 | hehe |
| 16:20.45 | brlcad | it's hard enough to find open source talent that are interested/willing to make things better |
| 16:20.53 | brlcad | even for something completely open like brl-cad |
| 16:21.22 | thing0 | yeah I bet |
| 16:21.24 | thing0 | I am still thinking of what todo with my free time |
| 16:21.27 | thing0 | not sure what yet |
| 16:21.43 | brlcad | there are reasons for that too, even not counting the mathematics, algorithms, and programming expertise generally needed |
| 16:22.17 | brlcad | it's just a big task, hard to get momentum going on something that big |
| 16:22.53 | thing0 | yep |
| 16:23.24 | thing0 | are you the project manager brlcad? |
| 16:23.40 | brlcad | yes |
| 16:23.56 | thing0 | ic |
| 16:24.15 | thing0 | well |
| 16:24.26 | brlcad | one thing that will hopefully spark activity that I'm looking to get in place this year is setting up an on-going GSoC-style program |
| 16:24.48 | thing0 | Gsoc? |
| 16:25.01 | brlcad | ~gsoc |
| 16:25.03 | ibot | gsoc is, like, the Google Summer of Code, a program run annually by Google to provide (paid for) jobs to students to code on open source projects over summer. See http://code.google.com/soc/ for details. |
| 16:25.16 | thing0 | yeah |
| 16:25.28 | thing0 | i didn't know they had an acroynm for it |
| 16:25.40 | thing0 | I forgot to apply last year |
| 16:25.45 | thing0 | missed the date |
| 16:25.50 | thing0 | I was PISSED ;) |
| 16:25.54 | brlcad | :) |
| 16:26.17 | brlcad | i missed the (mentorship) submission deadline both the first two years by mere days |
| 16:26.19 | thing0 | one of the things I am looking at is trying to get some PM skills |
| 16:26.28 | thing0 | I am a project engineer at the moment |
| 16:26.30 | brlcad | made it this year, got bzflag accepted |
| 16:26.36 | thing0 | but I want to be a PM eventually |
| 16:26.56 | thing0 | was trying to find something that I am interested in to try to develop PM skills |
| 16:26.57 | brlcad | titles are a bit superfluous ;) |
| 16:27.06 | thing0 | so that I can become a better PM |
| 16:27.10 | thing0 | yeah I know |
| 16:27.26 | thing0 | but, its the capabilities which is what I want |
| 16:27.58 | thing0 | i have spent many a night doing work |
| 16:28.10 | thing0 | but if I could harness other people |
| 16:28.13 | thing0 | could get more done |
| 16:28.22 | thing0 | still refining the skill of delegation |
| 16:28.35 | thing0 | need to be able to create understandable tasks |
| 16:28.36 | brlcad | my roles fluctuate between pm, architect, developer, code reviewer, tester, interface designer, graphics artist, etc .. even though each is really a "role" in itself |
| 16:28.38 | thing0 | :P\ |
| 16:28.46 | thing0 | hehe |
| 16:28.55 | thing0 | brlcad is the one man army |
| 16:28.56 | thing0 | hehe |
| 16:29.27 | brlcad | yeah, harnessing others and attracting development interest is one of the hardest parts |
| 16:29.31 | brlcad | you need a foundation for that |
| 16:29.33 | brlcad | both in the project |
| 16:29.40 | brlcad | and in your personality/goals/direction |
| 16:30.08 | thing0 | yep |
| 16:30.43 | thing0 | have you got milestones set for this year? |
| 16:31.38 | brlcad | for what it's worth, it's my belief that you have to not only be willing to do the work yourself that you ask of others, but that you also have the experience of having done the work you ask of them sometime before (i.e. base technical experience) |
| 16:31.56 | thing0 | exactly |
| 16:31.58 | brlcad | yeah, there are milestones as well as a rolling log |
| 16:32.14 | thing0 | because if u haven't done it yourself |
| 16:32.15 | brlcad | the brep steps I already mentioned |
| 16:32.21 | thing0 | how do you know if they did it right? |
| 16:32.35 | brlcad | there are other items documented in the TODO transcript, at least at a low-level |
| 16:32.48 | thing0 | i see |
| 16:33.00 | thing0 | is there a mailing list? |
| 16:33.07 | brlcad | not so much that they've done it right/wrong -- there are varying degrees of both in any implementation |
| 16:33.13 | brlcad | there are four mailing lists |
| 16:33.18 | brlcad | for different purposes |
| 16:33.41 | brlcad | http://sourceforge.net/mail/?group_id=105292users |
| 16:33.46 | brlcad | oops, http://sourceforge.net/mail/?group_id=105292 |
| 16:33.56 | brlcad | er, and that's five |
| 16:34.00 | thing0 | <PROTECTED> |
| 16:35.14 | brlcad | whether they took the path you wanted or not isn't really the issue, it's empowering them in directions that you know are good to be going in and knowing what those directions are or need to be |
| 16:35.42 | thing0 | hmm |
| 16:37.13 | brlcad | regardless of the "process" for getting them there or any given implementation approach sometimes |
| 16:39.30 | thing0 | well brlcad you have given me some stuff to think about |
| 16:39.58 | brlcad | thing0: for what it's worth, the task of a cad package is big enough to have entirely independent projects going -- if you wanted to exercise your project engineering or project management skills, theres undoubtely an area you could work on |
| 16:41.10 | brlcad | it's generally set up as a meritocracy which basically means participation and involvement (at any level, whether coding or administrative or otherwise) are what dictate direction, voice, and decisions |
| 16:41.39 | thing0 | hmm |
| 16:42.21 | thing0 | I just don't want to commit to something unless I am going to stick with it |
| 16:42.32 | thing0 | I hate doing that ;) |
| 16:46.41 | brlcad | just saying the participation door is wide open :) |
| 16:47.35 | thing0 | hehe |
| 16:47.40 | thing0 | thanks |
| 16:47.44 | thing0 | :) |
| 16:50.15 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/shaders.sh: use a shaders.mged transcript so that error reports are more concise |
| 16:54.42 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/Makefile.am: clean up after solids and shaders |
| 16:55.50 | thing0 | brlcad I take that you are Christopher Morrison |
| 16:59.03 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/solids.sh: remove solids.mged |
| 16:59.48 | thing0 | in regards to Gsoc would google be willing with them having sketchup |
| 16:59.50 | thing0 | ? |
| 17:02.24 | brlcad | thing0: brl-cad may or may not make gsoc itself, but we can certainly host our own program in a similar style |
| 17:03.06 | thing0 | ahh ic |
| 17:03.09 | brlcad | gsoc also has several limitations of its own that we wouldn't have to impose like only summer and only code |
| 17:03.22 | thing0 | lol |
| 17:03.50 | brlcad | we could do a brl-cad winter of code for example, or even year-round running once a quarter or something similar |
| 17:04.04 | thing0 | yeah |
| 17:04.22 | brlcad | there are interested groups that are willing/interested in sponsoring such a setup for brl-cad |
| 17:04.31 | brlcad | and yes, i'm morrison |
| 17:04.37 | thing0 | ic |
| 17:04.49 | thing0 | this could be interesting |
| 17:05.12 | brlcad | could be very interesting |
| 17:05.24 | thing0 | hehe |
| 17:06.52 | brlcad | for a mere staff year's funding, we could fund about 20 students for a gsoc-level timeframe (about three months) -- that's way more than we could handle right now managerially and with appropriate code review and mentoring |
| 17:07.49 | thing0 | wow |
| 17:07.50 | brlcad | even having just a handful students would accellerate interest and activity exceptionally |
| 17:08.02 | thing0 | yeah I can see that |
| 17:08.14 | thing0 | I think we could do recruitment drives |
| 17:08.30 | thing0 | to multidisplinary engineers |
| 17:08.34 | brlcad | yep, or even just "task drives" to get things done |
| 17:08.35 | thing0 | not just software |
| 17:08.41 | brlcad | ala month of bugs style stuff |
| 17:09.06 | thing0 | I am thinking Mechatronic (not just because I am one....) but because of the mechanical/software interest |
| 17:09.24 | thing0 | or mech eng who are doing com sci |
| 17:09.48 | thing0 | it is interesting that you have funding |
| 17:09.58 | thing0 | the first problem is usually funding |
| 17:10.21 | thing0 | but now that, that is sorted the next is people |
| 17:11.01 | brlcad | yeah, getting the mech-e, drafters, finite element analysts, solid modelers, elec-e, and others all working together on tool(s) that we all want and need |
| 17:11.49 | thing0 | yeah |
| 17:11.50 | thing0 | oh |
| 17:12.02 | thing0 | not trying to distract from the current goal |
| 17:12.09 | thing0 | but to through it into the mix |
| 17:12.17 | brlcad | many of the goals coincide though |
| 17:12.22 | thing0 | functional based modelling |
| 17:12.23 | brlcad | it's just a massive domain |
| 17:12.55 | brlcad | akin to the diagram I put together .. it's generally all "CAD", but the disciplines are vast with major overlap in some areas |
| 17:13.13 | thing0 | using languages like modelica to describe the geometry |
| 17:13.20 | brlcad | though many of the disciplines have different needs and expectations |
| 17:13.20 | thing0 | in a procedural like language |
| 17:14.24 | thing0 | yeah I remember reading that paper by that guy from lockhead |
| 17:14.58 | thing0 | let me find the reference |
| 17:16.40 | thing0 | http://www.voughtaircraft.com/ntcoe/presentations.htm |
| 17:18.20 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/ (Makefile.am gqa.sh spdi.sh weight.sh): move more of the mged scripts to their own files so that segfault failures on linux are more brief. |
| 17:19.22 | brlcad | heh, the lockhead guy is apparently a lisp/scheme fan :) |
| 17:19.29 | thing0 | yeah |
| 17:19.44 | thing0 | brlcad: have you ever seen used modelica? |
| 17:19.51 | brlcad | ah, haskell more specifically |
| 17:19.54 | thing0 | *seen/used |
| 17:20.23 | brlcad | his case points are a huge reason why brl-cad was open sourced in the first place |
| 17:21.04 | thing0 | exactly |
| 17:21.15 | thing0 | I HATE properitary formats |
| 17:21.20 | thing0 | they are so painful |
| 17:21.45 | brlcad | control, long-term accessibility, implementation openness, cross-platform identical behavior, NOT a proprietary binary format, uninhibited license, etc |
| 17:22.28 | thing0 | yep |
| 17:22.44 | brlcad | "that's why we're here" :) |
| 17:23.41 | thing0 | hehe |
| 17:23.42 | thing0 | yeah |
| 17:23.52 | brlcad | those ideas for a functional language ascii file format are interesting -- there are certainly some tradeoffs but the base idea is good |
| 17:24.21 | thing0 | PTC engineering for there software has had huge chunks out sourced to India |
| 17:24.31 | thing0 | this may seem unrelated |
| 17:24.33 | thing0 | but I was thinking |
| 17:24.34 | brlcad | it's actually not far from brl-cad existing ascii file format, though we're intently procedural via the tcl subset -- you're not going to get functional without a functional language :) |
| 17:24.47 | thing0 | it is because of the prohibitably costs |
| 17:24.55 | thing0 | that CAD vendors are charging |
| 17:25.12 | brlcad | they charge that much because they can :) |
| 17:25.18 | thing0 | exactly |
| 17:25.30 | thing0 | but PTC has had a lot of competition lately |
| 17:25.35 | thing0 | so to cut costs |
| 17:25.40 | brlcad | brl-cad's the next closest competitor and our gui is painful enough that they're not worried about us taking any business anytime soon |
| 17:25.44 | thing0 | design goes to India |
| 17:26.03 | brlcad | next closest open source competitor that is |
| 17:26.16 | thing0 | ?? |
| 17:26.30 | thing0 | who is the next closest open source competitor? |
| 17:27.13 | brlcad | brl-cad is really the only open source package that is anywhere close to being a competitor to the commercial packages |
| 17:27.18 | brlcad | and we're still pretty far |
| 17:27.33 | thing0 | ahh ok |
| 17:28.12 | brlcad | i mean there are several projects that have a gui or mesh editing or maybe parametrics or some other subset, but all of that really just barely scratches the surface of everything you need to be a real competitor |
| 17:28.33 | brlcad | the landscape of what everyone wants really is massive, and takes a lot of effort |
| 17:28.47 | thing0 | yeah |
| 17:28.54 | brlcad | brl-cad's only as far along as it is because it's been under constant development for over two decades now |
| 17:29.31 | brlcad | we certainly have the capability to compete -- like I said, we have most of the foundation -- we mostly lack in gui |
| 17:30.22 | thing0 | hmm |
| 17:30.29 | thing0 | it's sorta funny |
| 17:30.32 | brlcad | last estimate I saw puts brl-cad at roughly 500 man-years invested |
| 17:30.44 | thing0 | nice |
| 17:30.50 | brlcad | you wouldn't think it looking at mged :) |
| 17:31.15 | thing0 | hehe |
| 17:31.15 | brlcad | until you knew it inside and out |
| 17:31.20 | thing0 | the GUI was the thing that annoyed me the most in current CAD systems |
| 17:31.31 | thing0 | I wanted more of a consitent UI |
| 17:31.37 | thing0 | no dialog boxes |
| 17:31.46 | thing0 | more sorta consoleish |
| 17:31.49 | thing0 | in the sense that |
| 17:32.23 | thing0 | that there should be an area where the current tool should appear |
| 17:32.27 | brlcad | then there's the problem of what is consistent to one engineer can be outright "repellant" to another ;) |
| 17:32.30 | thing0 | it should always be the same |
| 17:32.39 | brlcad | "current tool should appear"? |
| 17:32.46 | thing0 | yeah |
| 17:32.51 | thing0 | let me explain |
| 17:32.57 | thing0 | when I want todo an extrusion |
| 17:33.02 | thing0 | a dialog popups up |
| 17:33.10 | thing0 | where I have to fill in values |
| 17:33.19 | brlcad | mkey |
| 17:33.25 | thing0 | there should be a part of the screen that is dedicated to it |
| 17:33.35 | thing0 | solidworks sorta have the idea |
| 17:33.50 | thing0 | but it is not that well implemented in my opinion |
| 17:33.52 | brlcad | ah, you mean non-overlapping? |
| 17:33.56 | thing0 | still very clunkey |
| 17:34.05 | thing0 | yes, non overlapping |
| 17:34.16 | brlcad | yes, that's something I've had a strong pet peave about |
| 17:34.23 | brlcad | modalities and dialogs in general |
| 17:34.29 | thing0 | yes |
| 17:34.32 | thing0 | I want a system |
| 17:34.41 | thing0 | whereby the whole screen is the modal space |
| 17:34.58 | thing0 | move the cursor over the edge of the screen and the bar fades in |
| 17:35.08 | thing0 | just like in some IDEs |
| 17:35.20 | thing0 | MSVC 2003 and up |
| 17:35.27 | thing0 | it just makes more sense |
| 17:35.35 | brlcad | jason (the guy working on brep) has an impressive design that is entirely non-modal, non-overlapping -- very impressive, very new .. there's not much out there like it interface-wise |
| 17:35.41 | thing0 | why waste screen real estate on tool bars? |
| 17:35.55 | thing0 | woah |
| 17:35.58 | thing0 | sounds cool |
| 17:36.05 | thing0 | sounds really cool |
| 17:36.58 | brlcad | even has a functioning prototype demo, although the demo is mac-only (as it was just a mockup) |
| 17:37.17 | thing0 | ic |
| 17:37.34 | *** join/#brlcad akreal (n=ak@87.249.56.198) | |
| 17:39.46 | brlcad | the concept, however, is pretty sound |
| 17:40.04 | thing0 | yeah |
| 17:40.09 | brlcad | considering using at least a lot of his ideas for the new editor interface |
| 17:40.10 | thing0 | I am just thinking |
| 17:40.30 | thing0 | have you thought about what the toolkit will be for UI? |
| 17:41.18 | brlcad | here's some of his thoughts on the matter, though he's writing from the grander aspect of an operating system: http://www.imaginaryday.com/wiki/ |
| 17:41.38 | brlcad | the ideas hold for any "big system" though that has to manage a slew of data types and interactions |
| 17:41.47 | thing0 | yeah |
| 17:41.52 | brlcad | i have though a LOT about the toolkit for the UI |
| 17:41.57 | thing0 | have you seen the interface of UGS NX5? |
| 17:42.33 | brlcad | the problem is that there's not much that I'm happy with so I'd not settled on anything at the moment -- only the high-level requirements from an architecture and integration perspective |
| 17:42.45 | brlcad | yes, I have |
| 17:43.15 | brlcad | i've got screenshots of just about all of the major cad systems to compare and contrast ideas and high-level layout |
| 17:43.23 | thing0 | that is one thing that I always remember about UGS |
| 17:43.39 | thing0 | they claim todo actual studies of users interaction with the software |
| 17:43.41 | brlcad | adobe has been leading the way, though, in terms of gui and integration |
| 17:43.51 | thing0 | to try and make it better with every release |
| 17:44.01 | thing0 | oh really, adobe |
| 17:44.02 | thing0 | hmm |
| 17:44.04 | thing0 | cool |
| 17:45.24 | brlcad | yeah, I worked with UGS pre-NX directly for several weeks analyzing their gui and interaction methods -- even back then it wasn't too bad, better than most really -- but it was still a complex system that took extensive effort and training to get started |
| 17:46.19 | thing0 | yeah |
| 17:46.38 | thing0 | but it seems that all parametric cads are the same |
| 17:46.54 | thing0 | I don't seem to have any trouble moving around |
| 17:47.24 | thing0 | I went from Inventor to Solidworks to Inventor to UGS to CATIA to Inventor |
| 17:47.34 | thing0 | i mean |
| 17:47.42 | thing0 | it's all the same philosphy |
| 17:47.49 | brlcad | yeah, once you learn the domain, it's not so bad |
| 17:47.56 | thing0 | yeah |
| 17:47.57 | brlcad | the entry to that domain is heavy though |
| 17:48.01 | brlcad | very heavy |
| 17:48.08 | thing0 | that's the problem |
| 17:48.12 | thing0 | it doesn't need to be |
| 17:48.16 | brlcad | you've got a lot of expert knowledge and experience under your belt to be able to do that |
| 17:48.49 | thing0 | I remember a quote from a guy I taught, the domain seems to be counter-intuitive sometimes with a lot of idosyncrincies |
| 17:49.20 | thing0 | everything non inventor in my previous list was from curiosity |
| 17:49.40 | thing0 | I play with the competition so that I could have a better understanding of how the history of it all evolve |
| 17:49.41 | thing0 | d |
| 17:49.52 | thing0 | I don't want to come across as a person who knows it all |
| 17:49.54 | thing0 | cause I don't |
| 17:50.02 | thing0 | I am a tinker by trade :) |
| 17:50.23 | thing0 | although, I did use solidworks for a bit of a longer period |
| 17:52.55 | brlcad | sounds familiar ;) |
| 17:53.39 | thing0 | Both of the above waste my time (which I personally consider rather precious), distract me from my work |
| 17:53.43 | brlcad | you at least have an understanding and appreciation for how easy it can/should be and want it to be better, that's great |
| 17:53.47 | thing0 | OMG I have used that before |
| 17:54.39 | brlcad | so do you code as well, or mostly user-level interaction? |
| 17:55.06 | brlcad | sounds like you know how to code |
| 17:55.45 | thing0 | i taught myself C when I was bored |
| 17:56.02 | thing0 | before going to uni and sitting through a C class which I already knew |
| 17:56.05 | thing0 | :) |
| 17:56.38 | thing0 | it's a skill that I have used every now and then |
| 17:56.54 | thing0 | I am using it at the moment at my current job to help automate some tasks |
| 17:56.59 | thing0 | but I mean |
| 17:57.06 | thing0 | it is extremly simple |
| 17:57.16 | thing0 | I need to ramp up my programming skills a bit |
| 17:57.21 | thing0 | they are a tad dusty :D |
| 17:57.47 | brlcad | gotcha |
| 17:57.58 | thing0 | just had more important things ;) |
| 17:58.13 | thing0 | teaching CAD requires you to drive the program, not write it ;) |
| 17:59.10 | brlcad | yep |
| 17:59.39 | brlcad | whilst working on the docs for brl-cad, code trickles almost to a halt |
| 18:00.10 | brlcad | documentation is a full-time job in itself, and a critical one depending on the interface |
| 18:00.25 | thing0 | yeah I know what you mean |
| 18:00.47 | thing0 | I have had to be a heck of a lot more thoughtful in my little automation exercise at work |
| 18:00.59 | thing0 | I want it to be fully traceable |
| 18:01.15 | thing0 | cause the software we are using from our client is a bit sketchy |
| 18:17.27 | *** join/#brlcad Elperion (n=Bary@p54876A32.dip.t-dialin.net) | |
| 18:22.50 | *** join/#brlcad Waimea (i=clock@77-56-111-45.dclient.hispeed.ch) | |
| 18:30.25 | brlcad | there he be, howdy Waimea |
| 18:35.12 | akreal | hello everybody! |
| 18:35.57 | brlcad | hello akreal |
| 18:37.10 | thing0 | hey akreal |
| 18:38.34 | akreal | i see that you discussed toolkit for the GUI, what do you think about Qt? |
| 18:41.45 | thing0 | well |
| 18:41.45 | thing0 | I was looking at this before |
| 18:41.45 | thing0 | just recently |
| 18:41.45 | thing0 | Qt is quite good |
| 18:41.45 | thing0 | I have tinkered with it before |
| 18:41.50 | *** join/#brlcad MinstrelGypsy (n=mario_du@bas2-sudbury98-1177871953.dsl.bell.ca) [NETSPLIT VICTIM] | |
| 18:41.50 | *** join/#brlcad yukonbob (n=bch@whthyt247-240.northwestel.net) [NETSPLIT VICTIM] | |
| 18:41.50 | *** join/#brlcad bjorkBSD (n=bjork@ip70-178-214-102.ks.ks.cox.net) [NETSPLIT VICTIM] | |
| 18:42.40 | brlcad | Qt is very good, their biggest limitation (imho) is their half-free licensing (which makes using them political), perception of complexity, and limited cross-platform support (i.e. to more than just win, mac, linux) |
| 18:43.07 | brlcad | otherwise, they're probably the best toolkit right now, with lots over their biggest competitor, gtk |
| 18:43.44 | brlcad | that said, they also push the same general gui interface of windows, icons, desktops, etc with anything non-standard being just as hard if not harder than going all-out custom (gui-wise) |
| 18:45.06 | thing0 | hehe I was about to talk about the licensing issue, but I thought I would check incase something had changed since the last time I played with qt (a year or two back ) |
| 18:45.26 | brlcad | blender's approach of a gui fully designed in opengl is a great one (imho) as it is the most expressive for something as complex as this task domain, but lacks in other regards of consistency and integration -- that mainly being a limitation of their implementation itself, though, and not opengl guis in general |
| 18:45.47 | brlcad | thing0: it's bad enough that you had to check -- that's what causes the political charge |
| 18:47.36 | brlcad | if you look for opengl gui toolkits, though, there isn't much out there (that I've found) that is any good |
| 18:47.37 | brlcad | the best you can probably find popularity-wise is cegui |
| 18:47.37 | *** join/#brlcad MinstrelGypsy (n=mario_du@bas2-sudbury98-1177871953.dsl.bell.ca) [NETSPLIT VICTIM] | |
| 18:47.37 | *** join/#brlcad yukonbob (n=bch@whthyt247-240.northwestel.net) [NETSPLIT VICTIM] | |
| 18:47.37 | *** join/#brlcad bjorkBSD (n=bjork@ip70-178-214-102.ks.ks.cox.net) [NETSPLIT VICTIM] | |
| 18:49.31 | brlcad | i have a whole matrix of concerns written down in a spreadsheet somewhere that talks about all of the various weighted considerations and available approaches |
| 18:50.33 | brlcad | ah, here we go |
| 18:50.40 | brlcad | tk, cegui, gtk, qt, clangui, wxwidgets, agar, libgui, fltk, custom |
| 18:51.02 | smallfoot- | oh many |
| 18:51.11 | brlcad | that was just gui toolkits |
| 18:51.21 | smallfoot- | ya |
| 18:51.34 | brlcad | for windowing and context management, that was a completely separate consideration |
| 18:52.36 | thing0 | i used fltk |
| 18:52.38 | brlcad | qt, clanlib, osg, sdl, ogre, java, gtk, glut, wxwidgets, custom (in no particular order) |
| 18:52.43 | thing0 | it was quite good |
| 18:52.49 | thing0 | so simple and lightweight |
| 18:52.55 | thing0 | fully configurable |
| 18:53.02 | *** join/#brlcad yukonbob (n=bch@whthyt247-240.northwestel.net) [NETSPLIT VICTIM] | |
| 18:53.03 | *** join/#brlcad bjorkBSD (n=bjork@ip70-178-214-102.ks.ks.cox.net) [NETSPLIT VICTIM] | |
| 18:55.15 | *** join/#brlcad MinstrelGypsy (n=mario_du@bas2-sudbury98-1177871953.dsl.bell.ca) [NETSPLIT VICTIM] | |
| 18:55.48 | *** part/#brlcad MinstrelGypsy (n=mario_du@bas2-sudbury98-1177871953.dsl.bell.ca) | |
| 18:56.07 | brlcad | yeah, it's got lots of nice attributes and configurability |
| 18:56.16 | brlcad | even customization, ala http://www.fltk.org/applications/images/SpiralSynth.png |
| 18:56.23 | thing0 | yeah |
| 18:56.35 | thing0 | I was using it on a personal project |
| 18:56.57 | thing0 | so I spent at least a month or 2 using it |
| 18:57.04 | thing0 | was quite fun ;) |
| 18:57.05 | brlcad | their downsides are similar to cegui though |
| 18:57.30 | brlcad | fixed widgets iirc, not scalable -- and not drawn via opengl? (don't remember) |
| 18:58.02 | brlcad | making seamless integration with a 3d environment (widget-wise) a bit difficult/impossible |
| 18:58.34 | thing0 | yeah |
| 18:58.39 | thing0 | i mean |
| 18:58.43 | brlcad | and they do lose points on perception and popularity |
| 18:58.46 | thing0 | I cannot confirm nor deny that |
| 18:59.22 | thing0 | based off UI interface we were discussing before |
| 18:59.37 | thing0 | shouldn't we just be looking at a open gl front end |
| 18:59.46 | *** join/#brlcad RodGallowGlass (n=mario_du@bas2-sudbury98-1177871953.dsl.bell.ca) | |
| 19:00.22 | thing0 | hi RodGallowGlass |
| 19:00.32 | RodGallowGlass | hey |
| 19:00.48 | brlcad | that's why I broke it out into graphics context and gui separately since you can embed many of the toolkits in any context |
| 19:01.59 | brlcad | even for the UI discussed, that could still be implemented custom or via toolkits like cegui, libgui, or customized versions of those even |
| 19:02.16 | thing0 | ahh ok |
| 19:02.19 | brlcad | RodGallowGlass: glad to read you made some .1 progress :) |
| 19:02.24 | thing0 | I see what ur getting at |
| 19:02.29 | RodGallowGlass | :) |
| 19:02.40 | RodGallowGlass | forced, i hate forcing things ;) |
| 19:03.31 | brlcad | tis good progress though |
| 19:03.44 | RodGallowGlass | yeah, I'm happy |
| 19:03.45 | brlcad | and you already helped pin a couple issues down in .1 |
| 19:03.56 | RodGallowGlass | didn't mean to ;) |
| 19:05.58 | RodGallowGlass | could it be possible the x11 summary report is inverted, i mean it said no x11 but seems to be building agaionst it anyway. |
| 19:07.35 | RodGallowGlass | well while this churns away, i'm going to visit the garden, be back later |
| 19:18.15 | thing0 | brb |
| 20:07.20 | thing0 | i'm going now |
| 20:07.24 | thing0 | cya later people |
| 21:19.16 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1096782371.dsl.bell.ca) | |
| 21:26.19 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-067-161.pools.arcor-ip.net) | |
| 23:03.21 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/src/librt/db_open.c: don't dereference null if we're failing early |
| 23:18.09 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/regress/moss.sh: fix call to pixdiff and ws |
| 23:53.01 | RodGallowGlass | ;) |
